In Search of Exomoons † David M. Kipping∗ Harvard-Smithsonian Center for Astrophysics, Cambridge, MA, USA E-mail: [email protected] Two decades ago, astronomers began detecting planets orbiting stars other than our Sun, so-called exoplanets. Since that time, the rate of detections and the sensitivity to ever-smaller planets has improved dramatically with several Earth-sized planets now known. As our sensitivity dives into the terrestrial regime, increasingly the community has wondered if the moons of exoplanets may also be detectable, so-called “exomoons”. Their detection represents an outstanding challenge in modern astronomy and would provide deep insights into the uniqueness of our Solar System and perhaps even expand the definition of habitability. Here, I will briefly review theoretical studies exploring the formation and evolution of exomoons, which serve to guide observational searches and provide testable hypotheses. Next, I will outline the different methods which have been proposed to accomplish this challenging feat and their respective merits. Finally, initial results from observational efforts will be summarized with a view to future prospects as well. arXiv:1405.1455v1 [astro-ph.EP] 6 May 2014 Frank N. Bash Symposium 2013: New Horizons in Astronomy (BASH 2013) October 6-8, 2013 Austin, Texas ∗Speaker. †NASA Carl Sagan Fellow. c Copyright owned by the author(s) under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike Licence. http://pos.sissa.it/ In Search of Exomoons David M. Kipping 1. INTRODUCTION Within the field of astronomy, there are few areas of research which have enjoyed as much public enthusiasm and rapid rates of discovery than that of exoplanetary science - a research area which was not even established until two decades ago. Astronomers first waded out around the shores of this novel discipline around 1990, with compelling evidence for exoplanets found by [1], [2] and [3]. However, it was not until the mid-nineties, after the discovery of 51 Pegasi b by [4], that astronomers unfurled their sails and began regularly ensnaring impressive hauls of exoplanets. Since then, more than one thousand confirmed exoplanets and five thousand candidates have been detected. Furthermore, observational sensitivity has dramatically improved from being lim- ited to hot-Jupiters back in the mid-nineties [5] to now being able to detect Earth-sized habitable- zone worlds [6]. This progress stems from a wide range of improvements in the instrumentation, telescope time, use of dedicated missions, man-power, observing strategy, modeling and of course research funding, which enables all of the above [7]. This journey of scientific discovery is surely not complete though, and many mysteries and outstanding challenges remain. For example, we now know that our solar system and our planet represent just a drop in the cosmic ocean, and the surprising diversity of planetary architectures and compositions indicates that we lack the deep understanding of how planets form and evolve [8]. Similarly, whilst empirical estimates of the occurence rate of habitable-zone rocky planets have been made [9, 10], the detection of bio-signatures or detailed atmospheric characterization of such worlds is still some way off [11]. With our sensivity to exoplanets reaching the level of Earth and even sub-Earth sized worlds [6], one outstanding question which may be within reach is the detection of extrasolar moons; so-called “exomoons”. Their detection would likely usher in an entire new sub-field of scientific endeavor, much like the wind change instigated by the discovery of 51-Pegasi-b. Measuring the physical and orbital characteristics of ensemble populations of exomoons would likely yield power- ful constraints on the formation and evolution mechanisms guiding alien worlds [12]. Such bodies may themselves be habitable [13, 14, 15, 16] or affect the habitability of the planets which they orbit [17]. A confirmed exomoon detection would therefore represent a major scientific breakthrough, yet this lofty goal remains a daunting and outstanding challenge to modern astronomy. In this work, I will briefly discuss several plausible formation mechanisms for exomoons in §2, with a view as to how these ideas may impact observational strategies. In §3, I will review the proposed detection techniques and sensitivities, with focus on the transit-based methods. In §5, empirical results to date will be summarized and placed into context. Finally, in §6, we discuss future prospects for exomoon detection with current and planned exoplanet missions. 2 In Search of Exomoons David M. Kipping 2. FORMATION & EVOLUTION 2.1 Context We begin by briefly discussing several possible formation and subsequent evolution mecha- nisms for exomoons. Since we currently know of no confirmed exomoons, this topic is strongly influenced by inspection of the moons within our solar system, so-called “endomoons”. Given the enormous diversity of exoplanetary architectures discovered to date, which rarely resemble the So- lar System, it is important to not put too much faith in such theories. Nevertheless, they provide a useful starting point and at least offer hypotheses which can be subsequently tested with observa- tions. I therefore endeavor to provide only a brief introduction to give a flavor of how theory could guide observing strategy. In our solar system, we observe two broad classes of satellites, although there exists no widely accepted definition as to what divides these two classes. In this work, I define regular satellites to have formed in-situ from the circumplanetary disk and these tend to display nearly-coplanar and nearly-circular orbits. In contrast, irregular satellites are those which formed via some other mecha- nism, such as an impact or a capture, and these tend to exhibit highly inclined (including retrograde motion) and highly eccentric orbital paths. Regular satellites also tend to have more compact orbits than the irregular population. Naturally, these two populations have distinct histories and thus I will split the discussion of formation and evolution accordingly. 2.2 Formation of Regular Satellites Although considerable diversity exists within the literature for proposed formation mecha- nisms of the regular satellites, the fact that such satellites have compact and prograde orbits has led to the general consensus that they formed from a circumplanetary disk [18]. In general, formation models focus on the Jovian (belonging to Jupiter) and Kronian (belonging to Saturn) satellites, since these two planets host the largest number of regular satellites and have enjoyed a long history of detailed observations. In this framework, the primary challenge is to understand the differences between the number of major moons. One of the leading formation models comes from a series of papers developed by [19, 20, 21] and is known as the actively supplied gaseous accretion disk model. In this model, dust grains within a circumplanetary disk stick and grow to form satellitesimals, which then migrate via type I migration and are disposed within 105 yrs [22]. Continuous mass-infall form the protoplanetary ∼ disk maintains a peak circumplanetary disk density of 100 g cm 2, allowing new satellitesimals ∼ − to continuously grow. Once the planet has opened up a gap in the protoplanetary disk, the active supply halts and the circumplanetary disk rapidly diffuses in 103 yrs, thus “freezing” the remain- ∼ ing satellites in place. The final total satellite mass is therefore given by a balance between type I migration disposal and the repeated satellitesimal accretion rate, which [19] argue to be universally 4 of order (∑MS)/MP 10 . In this model, the Jovian/Kronian differences are proposed to be due ∼ − to presence of an inner cavity within the circum-Jovian disk, which Saturn was unable to form due its distinct mass and semi-major axis [23]. An alternative model comes from [18, 24] and [25]: the solids enhanced minimum mass model. Here, the planet hosts a two-component disk comprised of a dense inner sub-disk sur- rounded by a lower density outer disk. This results in a much longer satellite migration timescale 3 In Search of Exomoons David M. Kipping than the associated formation timescale. The model qualitatively describes the expected mass ra- tios, but does not provide definitive predictions– unlike the actively supplied disk accretion model. 2.3 Formation of Irregular Satellites In this Review, I consider irregular satellites to be those which initially formed not as a satellite to the planet in question, but somehow ended up so via a capture or impact event. Before this event, the satellite may have been a planet/dwarf planet itself, a Trojan, a satellite of another planet or a binary pair. For moons to form via a capture or impact, one clearly requires strong dynamical mixing leading to close encounters. The ultimate outcome of an encounter will chiefly depend upon how close these encounters are (the impact parameter) and the relative masses and velocities of the bodies involved. In this sense, one can consider impacts to be a special case of a capture, where the impact parameter of the event is less than one planetary radius. Giant planets must frequently migrate given the observed population of hot Jupiters, which certainly could not have formed in-situ [26]. During this inward migration, the giant planet may encounter terrestrial planets and in such encounters a capture could transpire. In all capture sce- narios, one requires the relative velocities of the two bodies to be below that of the planet’s escape velocity, if the putative satellite is to be captured. This generally requires a deceleration of the putative satellite, for which several mechanisms have been proposed. In the case of the Moon, our satellite’s composition, rotation and orientation support the so- called “giant impact hypothesis” between the primordial Earth and a Mars-sized body some 4.5 Gyr ago [27]. Whilst this hypothesis explains much of the available evidence, it is unclear how fre- quently such events may occur or how special the Moon may be.